As with most complex problems, the simple, easy-to-understand hypothesis that novel traits arise from new genetic mutations is wrong. Although scientists have known this for decades, an alternative explanation has yet to present itself. On page 1004 of this issue, Hu et al. ([ 1 ][1]) show that deciphering the mechanisms underlying the evolution of a new biological trait is intimately intertwined with our evolving understanding of the concept of homology and its role in developmental genetics. Darwin's postulate of descent with modification implies that all biological traits must have an evolutionary antecedent—that is, every trait is structurally related to a preceding one. This concept of homology among traits was developed more than a decade before Darwin's major works by paleontologist Richard Owen, who defined it as: “…the same organ in different animals under every variety of form and function.” This seems clear enough until you parse the language and ask what exactly is meant by “the same.” Owen expanded on his original definition by noting that homology among traits could be recognized by correspondences in their structure, position in the body, and embryonic development. Paradoxically, as more has been learned about the development of biological structures, it has become increasingly difficult to recognize homologies. Part of the reason is that most structures are built from preexisting parts with different developmental and evolutionary histories. This conundrum has led to the concepts of partial homology and homology of a selected attribute of a structure—but not the structure as a whole—with a feature in an ancestor ([ 2 ][2]–[ 4 ][3]). Modern developmental genetics has made the task of identifying homologies even more fraught by revealing that a surprisingly small toolbox of regulatory genes controls the development of exceptionally diverse and seemingly unrelated structures ([ 5 ][4]). Hu et al. investigated the developmental and genetic basis of an evolutionarily novel structure, the prothoracic horns of dung beetles (members of the family Scarabaeidae). The prothorax is the anteriormost of the three thoracic segments of insects. All three segments have a pair of ventral walking legs, and the second and third also bear a pair of dorsal wings. Among the great and continuing puzzles in insect evolution are the origin of the wings and why the prothorax does not have any. The dorsal prothorax has, however, undergone a tremendous diversification of structure in various lineages of insects, ranging from elaborate three-dimensional helmets in treehoppers; great crests in wheel bugs; crests and long spikes in some grasshoppers; and single, double, triple, or quadruple horns of scarabaeid beetles—a remarkable example of a secondary sexual trait in the animal kingdom. Wings and legs are thought to have evolved from a single common precursor that split into a dorsal part and a ventral part for the wing and leg, respectively. Moreover, recent work has shown that the wing also has a composite origin, deriving from parts of the dorsal and lateral plates of the thorax. This dual origin of the wing, it turns out, plays an important role in the developmental origin of horns. ![Figure][5] Dorsal wing field drives development of beetle horn Panels show the steps in the development of a horn in dung beetles. The horn develops at the larval stage, beginning in specified regions of tissue in the insect dorsal-wing development field. GRAPHIC: N. CARY/ SCIENCE , (PHOTOS, TOP TO BOTTOM) Y. HU ET AL. , ALEXANDER WILD Hu et al. began their elegant series of experiments by using RNA interference (RNAi) in scarabaeid beetle larvae and pupae to inhibit the expression of several genes known to be associated with wing development. They found that the beetle's wings were indeed reduced or lost completely. More surprising, however, was the observation that the large medial prothoracic horn was also reduced in size and split into two small bilateral projections. The authors then showed that vestigial , a selector gene for wing development, was expressed throughout the developing wings (as expected) but also at the base of the horn precursors in the pupal stage. Hu et al. also found that the medial prothoracic horn begins its development as two lateral protuberances during the pupal stage that gradually migrate to the middle of the pronotum. These findings led the authors to conclude that wing-related genes are required for initiating horn development and that, therefore, the prothoracic horns might share partial homology with wings. Hu et al. then used RNAi to inhibit expression of the Sex combs reduced ( Scr ) gene, whose normal activity specifies the prothoracic segment and in whose absence this segment acquires the characteristics of the second thoracic segment, which bears the forewings. This resulted in the development of structures that resembled the normal forewing (elytra), as well as a severe reduction in the size of the prothoracic horn, which was again reduced to two small lateral protuberances. Not all transformations of the thorax were equally severe. In fact, Hu et al. found a reciprocal relationship between the size of the wings and the size of the two horns that developed when the expression of individual wing-related genes was inhibited by RNAi; large wings appeared to develop at the expense of large horns and vice versa. Inhibiting expression of the pannier ( pnr ) gene, which specifies the medial region of the notum (the dorsal part of the insect's thoracic segment), confirmed the hypothesis that thoracic horn primordia provide the material from which the ectopic wings developed. Hu et al. then conducted a series of RNA-sequencing studies to determine what kinds of genes were expressed in the developing horns and adjacent tissues. They found that soon after horn development started, wing-specific gene expression ceased and further development of the horns was accompanied by new gene expression patterns. The overall conclusion is that horns arise from tissues that would have formed the dorsal component of the wing and use part of the wing's gene network to get started. However, subsequent growth and morphogenesis of the horn use a diverse set of genes unrelated to wing development (see the figure). The single medial horn of these beetles thus appears to come about by the migration and fusion of two lateral primordia derived from the dorsal wing field. It seems possible, therefore, that beetles that have two horns fail to merge their primordia, and those with four horns split the two primordia. The flexibility of using wing genes to initiate a primordium and then building on that foundation by selecting for genetic networks that promote specific morphogenesis might well have been the foundation for the elaborate and sometimes riotous diversification of form seen in the pronotum of insects. This raises a new question: What is the primitive function of what is now called the wing gene network? Is it to specify wings, or are wings latecomers in a network that evolved to regulate various outgrowths on the thorax and abdomen? Novelty may be in the eye of the beholder. 1. [↵][6]1. Y. Hu, 2. D. M. Linz, 3. A. P. Moczek , Science 366, 1004 (2019). [OpenUrl][7][Abstract/FREE Full Text][8] 2. [↵][9]1. V. L. Roth , Biol. J. Linn. Soc. Lond. 22, 13–29 (1984). [OpenUrl][10][CrossRef][11][Web of Science][12] 3. 1. M. J. West-Eberhard , Developmental Plasticity and Evolution (Oxford Univ. Press, 2003). 4. [↵][13]1. G. P. Wagner , Homology, Genes, and Evolutionary Innovation (Princeton Univ. Press, 2014). 5. [↵][14]1. S. B. Carroll, 2. J. K. Grenier, 3. S. D. Weatherbee , From DNA to Diversity: Molecular Genetics and the Evolution of Animal Design (Blackwell Science, 2004). [1]: #ref-1 [2]: #ref-2 [3]: #ref-4 [4]: #ref-5 [5]: pending:yes [6]: #xref-ref-1-1 "View reference 1 in text" [7]: {openurl}?query=rft.jtitle%253DScience%26rft.stitle%253DScience%26rft.aulast%253DHu%26rft.auinit1%253DY.%26rft.volume%253D366%26rft.issue%253D6468%26rft.spage%253D1004%26rft.epage%253D1007%26rft.atitle%253DBeetle%2Bhorns%2Bevolved%2Bfrom%2Bwing%2Bserial%2Bhomologs%26rft_id%253Dinfo%253Adoi%252F10.1126%252Fscience.aaw2980%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [8]: /lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6Mzoic2NpIjtzOjU6InJlc2lkIjtzOjEzOiIzNjYvNjQ2OC8xMDA0IjtzOjQ6ImF0b20iO3M6MjI6Ii9zY2kvMzY2LzY0NjgvOTQ2LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ== [9]: #xref-ref-2-1 "View reference 2 in text" [10]: {openurl}?query=rft.jtitle%253DBiol.%2BJ.%2BLinn.%2BSoc.%2BLond.%26rft.volume%253D22%26rft.spage%253D13%26rft_id%253Dinfo%253Adoi%252F10.1111%252Fj.1095-8312.1984.tb00796.x%26rft.genre%253Darticle%26rft_val_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Ajournal%26ctx_ver%253DZ39.88-2004%26url_ver%253DZ39.88-2004%26url_ctx_fmt%253Dinfo%253Aofi%252Ffmt%253Akev%253Amtx%253Actx [11]: /lookup/external-ref?access_num=10.1111/j.1095-8312.1984.tb00796.x&link_type=DOI [12]: /lookup/external-ref?access_num=A1984SV01300002&link_type=ISI [13]: #xref-ref-4-1 "View reference 4 in text" [14]: #xref-ref-5-1 "View reference 5 in text"